Social Psychiatry and Psychiatric Epidemiology

, Volume 50, Issue 3, pp 479–487

The utility of estimating population-level trajectories of terminal wellbeing decline within a growth mixture modelling framework

  • R. A. Burns
  • J. Byles
  • D. J. Magliano
  • P. Mitchell
  • K. J. Anstey
Original Paper

DOI: 10.1007/s00127-014-0948-3

Cite this article as:
Burns, R.A., Byles, J., Magliano, D.J. et al. Soc Psychiatry Psychiatr Epidemiol (2015) 50: 479. doi:10.1007/s00127-014-0948-3



Mortality-related decline has been identified across multiple domains of human functioning, including mental health and wellbeing. The current study utilised a growth mixture modelling framework to establish whether a single population-level trajectory best describes mortality-related changes in both wellbeing and mental health, or whether subpopulations report quite different mortality-related changes.


Participants were older-aged (M = 69.59 years; SD = 8.08 years) deceased females (N = 1,862) from the dynamic analyses to optimise ageing (DYNOPTA) project. Growth mixture models analysed participants’ responses on measures of mental health and wellbeing for up to 16 years from death.


Multi-level models confirmed overall terminal decline and terminal drop in both mental health and wellbeing. However, modelling data from the same participants within a latent class growth mixture framework indicated that most participants reported stability in mental health (90.3 %) and wellbeing (89.0 %) in the years preceding death.


Whilst confirming other population-level analyses which support terminal decline and drop hypotheses in both mental health and wellbeing, we subsequently identified that most of this effect is driven by a small, but significant minority of the population. Instead, most individuals report stable levels of mental health and wellbeing in the years preceding death.


Well-being Mental health Mortality Epidemiology Mixture modelling 

Copyright information

© Springer-Verlag Berlin Heidelberg 2014

Authors and Affiliations

  • R. A. Burns
    • 1
  • J. Byles
    • 2
  • D. J. Magliano
    • 3
  • P. Mitchell
    • 4
  • K. J. Anstey
    • 1
  1. 1.Centre for Research on Ageing, Health and WellbeingThe Australian National UniversityCanberraAustralia
  2. 2.Research Centre for Gender, Health and AgeingUniversity of NewcastleNewcastleAustralia
  3. 3.BakerIDI Heart and Diabetes InstituteMelbourneAustralia
  4. 4.Centre for Vision Research, Westmead Millennium Institute and Department of Ophthalmologythe University of SydneySydneyAustralia

Personalised recommendations